Support and News Forum

Hi,

I'm using the latest EVE, but like in unetlab, I cannot use any Juniper device like vSRX or vMX.
When I start a Juniper node, I can see the qemu process in Linux terminal (top command), the console connects but nothing is displayed.
When I stop the juniper node, it shows am icon of an exclamation (!) inside a triangle in the place of the stop or start icon.
I cannot start again the juniper node, until I made a wipe on it.
I'm running the EVE VM in a VMware Workstation 12. My desktop is a AMD Phaenom II X6 1055t, ASUS M3A99X, 32Gb DDR3, Windows 7 x64.

I can run the vSRX or vMX perfectly in GNS3 1.5.3.
Also, in my troubleshooting, I made it run in previous unetlab, but i need to install the KVM inside the unetlab VM.
In GNS3 VM, KVM is already installed just for the comparison.

Can someone from EVE Team fix it? Thanks in advance!

Edit node, change config from Exported to None and way you go.

When you start node and see clock icon - that mean it will try to boot from exported config, you need to give it fully load before you can stop/work with it.
If you stop node during that process - it will show the triangle with ! and only wipe can fix that

Also, in my troubleshooting, I made it run in previous unetlab, but i need to install the KVM inside the unetlab VM. - nonsense, everything work in EVE out of the box

mike wrote: ↑

Tue Mar 28, 2017 5:51 pm

Edit node, change config from Exported to None and way you go.

Hi Mike!

I followed your answer, the triangle with !, disappeared when I stop the Juniper node. But the issue with the console, continues.
I tried with the embedded client in EVE (guacamole), no luck, it connect but nothing is displayed too.

What can be? I can start normally IOU, IOSv, etc.

Thanks in advance.

It is definitely image issue, or your PC CPU cannot run such node

Do you have any other qemu image running? If not, you need to fix nested virtualization issue

@mike
Yes, I have IOSv and IOSvL2 qemu images running fine in EVE.

@UD
It isn't an image problem, because I can run the same images very fine in GNS3. The vSRX image is original, I downloaded it from Juniper.

I think that is something like this user experienced: http://forum.802101.com/post1369.html#p1369

I can use GNS3 as well, but I would like to use EVE, because I can share my Labs here and in cert collection. I see people sharing only Cisco Labs and nothing about Juniper...

To be clear, the SecureCRT or Putty, stay connected to the node console port, but nothing is displayed and the cpu utilization of the qemu node is 100%...

which exacly image??
version?

Images and versions:
vSRX = junos-vsrx-12.1X47-D20.7-domestic.qcow2 (downloaded from Juniper)
vMX = jinstall-vmx-14.1R4.8-domestic.img

Also, I made a recent test using vQFX, it woks in my GNS3 but no luck in EVE. The PFE starts and close after some seconds, the RE qemu process still running at 100% of cpu and no output from console.
Images and versions:
vQFX =RE vqfx10k-re-15.1X53-D60.vmdk and PFE vqfx10k-pfe-20160609-2.vmdk

The EVE VM and GNS3 VM are very close to each other, both runs Ubuntu (EVE = Ubuntu 16.04.2 LTS) (GNS3 = 14.04.5 LTS), but EVE runs a QEMU version 2.4.0 and GNS3 1.5.3 runs QEMU version 2.5.0 compiled for Debian.
GNS3 VM support the kvm-ok linux command and shows KVM acceleration can be used and EVE don't support this command.
I'm not a linux expert, but it relies to something with EVE's QEMU version.

Sorry for this, but I cannot guess what is causing this issue...